通过之前解的一个bug,在通话过程中要切换Audio Output Path从蓝牙耳机到Speaker,但是却第一次却切换到了earpiece,再切换一次
才切到Speaker,我就根据这个bug的分析,来熟悉下音频通道的切换过程;
首先还是需要看下google对于android audio系统架构分析;
http://source.android.com/devices/audio.html
android的HAL(Hardware Abstraction Layer)是连接上层的audio-specific framework APIs和底层的audio驱动和硬体的关键;
下面这张图展示了这样的架构以及每层的代码分布:
一、Application framework
在Application framework层级是app层的code,是通过android.media提供的API来与audio硬件进行交互动作,这部分的代码是通过
audio JNI来调用native代码从而达到影响硬件的效果;
二、JNI
JNI部分的代码是位于 frameworks/base/core/jni/和frameworks/base/media/jni 目录下的;
三、Native framework
Native framework定义在 frameworks/av/media/libmedia里,它与android.media 包对应并且通过IBinder的IPC代理来访问底层
audio服务;
四、Binder IPC
Binder IPC通信是跨进程通信的手段,audio的这部分代码位于frameworks/av/media/libmedia目录下,并且命名都是以I开头的;
五、Media Server
Audio Service是隶属Media Server的,其代码位于 frameworks/av/services/audioflinger,它是真正的与HAL层的实现进行交互的;
六、HAL
HAL层定义了Audio Service调用的标准接口,不同的硬件必须根据自己的情况来实现这个接口来让硬件在android中正常的工作,
所以可以在不影响应用层系统调用的情况下,更换不同的硬件。大大减少了系统耦合性;
七、Kernel Driver
Audio驱动是与硬件进行交互,并且实现HAL层的接口供上层正常调用,这里,厂商可以选择ALSA,OSS以及自定义的音频驱动;
(NOTE:如果选择ALSA,android建议使用 external/tinyalsa目录下的实现);
接下来就来说说通话时音频通道的切换,但是往下看之前必须知道,对于Audio Path的切换,android有一策略管理器来帮我们分配好
输入输出的设备,比如当手机播放音乐时,从Speaker播放出来,这时候插入耳机的话会从耳机设备输出;但是有时候我们想要自己去
指定的话,就是我们接下来要说的了;
我们在通话时,要是开免提,实际上也就是Audio Path切换到了Speaker,也就是外方喇叭;代码中的话调用一个函数即可,这是强制
切换audio Path,不遵从系统的分配:
|
AudioManager audioManager = (AudioManager) context.getSystemService(Context.AUDIO_SERVICE);
audioManager.setSpeakerphoneOn(true);
|
中间过程简单不说,最终是调用到了JNI,android_media_AudioSystem中的android_media_AudioSystem_setForceUse()函数,
来看下其具体实现:
|
static int
android_media_AudioSystem_setForceUse(JNIEnv *env, jobject thiz, jint usage, jint config)
{
SLOGE("jni android_media_AudioSystem_setForceUse()");
return check_AudioSystem_Command(AudioSystem::setForceUse(static_cast <audio_policy_force_use_t>(usage),
static_cast <audio_policy_forced_cfg_t>(config)));
}
|
显而易见,它是调用了AudioSystem.cpp的setForceUse()函数,check_AudioSystem_Command()不说,重点看看
audio_policy_force_use_t和audio_policy_forced_cfg_t这两个结构体;audio_policy_force_use_t说明的是当前的Audio环境,
audio_policy_forced_cfg_t表示audio的输入输出设备;它们是专门为setForceUse所用的;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
|
/* usages used for audio_policy->set_force_use() */
typedef enum {
//表示的是通话过程中
AUDIO_POLICY_FORCE_FOR_COMMUNICATION,
//媒体
AUDIO_POLICY_FORCE_FOR_MEDIA,
//录音
AUDIO_POLICY_FORCE_FOR_RECORD,
AUDIO_POLICY_FORCE_FOR_DOCK,
AUDIO_POLICY_FORCE_FOR_SYSTEM,
AUDIO_POLICY_FORCE_USE_CNT,
AUDIO_POLICY_FORCE_USE_MAX = AUDIO_POLICY_FORCE_USE_CNT - 1,
} audio_policy_force_use_t;
|
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
/* device categories used for audio_policy->set_force_use() */
typedef enum {
AUDIO_POLICY_FORCE_NONE,
AUDIO_POLICY_FORCE_SPEAKER,
AUDIO_POLICY_FORCE_HEADPHONES,
AUDIO_POLICY_FORCE_BT_SCO,
AUDIO_POLICY_FORCE_BT_A2DP,
AUDIO_POLICY_FORCE_WIRED_ACCESSORY,
AUDIO_POLICY_FORCE_BT_CAR_DOCK,
AUDIO_POLICY_FORCE_BT_DESK_DOCK,
AUDIO_POLICY_FORCE_ANALOG_DOCK,
AUDIO_POLICY_FORCE_DIGITAL_DOCK,
AUDIO_POLICY_FORCE_NO_BT_A2DP, /* A2DP sink is not preferred to speaker or wired HS */
AUDIO_POLICY_FORCE_SYSTEM_ENFORCED,
AUDIO_POLICY_FORCE_CFG_CNT,
AUDIO_POLICY_FORCE_CFG_MAX = AUDIO_POLICY_FORCE_CFG_CNT - 1,
AUDIO_POLICY_FORCE_DEFAULT = AUDIO_POLICY_FORCE_NONE,
} audio_policy_forced_cfg_t;
|
这时候我们就应该知道,当我想要在通话时打开Speaker,传递的参数就是usage和config分别是
AUDIO_POLICY_FORCE_FOR_COMMUNICATION和AUDIO_POLICY_FORCE_SPEAKER了,这两个参数从上层一直到底层,
还是很简单的;
接着往下看就是调用的AudioSystem.cpp的setForceUse()函数了;
|
status_t AudioSystem::setForceUse(audio_policy_force_use_t usage, audio_policy_forced_cfg_t config)
{
SLOGE("setForceUse() usage = %d, config = %d" ,usage , config);
const sp<IAudioPolicyService>& aps = AudioSystem::get_audio_policy_service();
if (aps == 0) return PERMISSION_DENIED;
return aps->setForceUse(usage, config);
}
|
get_audio_policy_service()函数不做过多解释,就是通过Native的ServiceManager来获取audio policy的Service代理对象,
从而实现与audio policy的进程间通讯;
|
.......
binder = sm->getService(String16("media.audio_policy"));
.......
|
接下来就是调用frameworks/av/services/audioflinger/AudioPolicyService.cpp的setForceUse()函数了;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
|
status_t AudioPolicyService::setForceUse(audio_policy_force_use_t usage,
audio_policy_forced_cfg_t config)
{
if (mpAudioPolicy == NULL) {
return NO_INIT;
}
if (!settingsAllowed()) {
return PERMISSION_DENIED;
}
if (usage < 0 || usage >= AUDIO_POLICY_FORCE_USE_CNT) {
return BAD_VALUE;
}
if (config < 0 || config >= AUDIO_POLICY_FORCE_CFG_CNT) {
return BAD_VALUE;
}
Mutex::Autolock _l(mLock);
mpAudioPolicy->set_force_use(mpAudioPolicy, usage, config);
return NO_ERROR;
}
|
这个mpAudioPolicy是什么呢?它的set_force_use函数在哪里实现呢?这两个问题需要了解就OK了;
首先mpAudioPolicy它是一个指针,在AudioServicePolicy.cpp的构造函数中被赋值,来看看其赋值过程:
|
......
const struct hw_module_t *module;
......
rc = hw_get_module(AUDIO_POLICY_HARDWARE_MODULE_ID, &module);
......
rc = audio_policy_dev_open(module, &mpAudioPolicyDev);
......
rc = mpAudioPolicyDev->create_audio_policy(mpAudioPolicyDev,&aps_ops,this, &mpAudioPolicy);
......
|
首先AUDIO_POLICY_HARDWARE_MODULE_ID值是:
|
#define AUDIO_POLICY_HARDWARE_MODULE_ID "audio_policy"
|
其次module是一个指针,指向的是一个hw_module_t结构体类型,它的作用是调用系统的哪个audio policy module,
这个module可以是原始的,也可以由厂商自定义的;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
|
typedef struct hw_module_t {
/** tag must be initialized to HARDWARE_MODULE_TAG */
uint32_t tag;
uint16_t module_api_version;
#define version_major module_api_version
uint16_t hal_api_version;
#define version_minor hal_api_version
/** Identifier of module */
const char *id;
/** Name of this module */
const char *name;
/** Author/owner/implementor of the module */
const char *author;
/** Modules methods */
struct hw_module_methods_t* methods;
/** module's dso */
void* dso;
/** padding to 128 bytes, reserved for future use */
uint32_t reserved[32-7];
} hw_module_t;
|
再来看看是如何给module赋值的;
|
hardware.c
int hw_get_module(const char *id, const struct hw_module_t **module)
{
return hw_get_module_by_class(id, NULL, module);
}
|
看看hw_get_module_by_class方法的实现:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
|
hardware.c
int hw_get_module_by_class(const char *class_id, const char *inst,
const struct hw_module_t **module)
{
int status;
int i;
const struct hw_module_t *hmi = NULL;
char prop[PATH_MAX];
char path[PATH_MAX];
char name[PATH_MAX];
if (inst)
snprintf(name, PATH_MAX, "%s.%s", class_id, inst);
else
strlcpy(name, class_id, PATH_MAX);
/*
* Here we rely on the fact that calling dlopen multiple times on
* the same .so will simply increment a refcount (and not load
* a new copy of the library).
* We also assume that dlopen() is thread-safe.
*/
/* Loop through the configuration variants looking for a module */
for (i=0 ; i<HAL_VARIANT_KEYS_COUNT+1 ; i++) {
if (i < HAL_VARIANT_KEYS_COUNT) {
if (property_get(variant_keys[i], prop, NULL) == 0) {
continue;
}
snprintf(path, sizeof(path), "%s/%s.%s.so",
HAL_LIBRARY_PATH2, name, prop);
if (access(path, R_OK) == 0) break;
snprintf(path, sizeof(path), "%s/%s.%s.so",
HAL_LIBRARY_PATH1, name, prop);
if (access(path, R_OK) == 0) break;
} else {
snprintf(path, sizeof(path), "%s/%s.default.so",
HAL_LIBRARY_PATH1, name);
if (access(path, R_OK) == 0) break;
}
}
status = -ENOENT;
if (i < HAL_VARIANT_KEYS_COUNT+1) {
/* load the module, if this fails, we're doomed, and we should not try
* to load a different variant. */
status = load(class_id, path, module);
}
return status;
}
|
方法是找到指定的库文件并且加载;不做详细介绍;这里会得到audio_policy.default.so;这个库正是编译
hardware/libhardware_legacy/audio出来的;
再跳回到AudioPolicyService的构造函数中来;接下来
|
rc = audio_policy_dev_open(module, &mpAudioPolicyDev);
|
它调用的是legacy_ap_dev_open()函数,不做详细介绍:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
|
audio_policy_hal.cpp
static int legacy_ap_dev_open(const hw_module_t* module, const char* name,
hw_device_t** device)
{
struct legacy_ap_device *dev;
if (strcmp(name, AUDIO_POLICY_INTERFACE) != 0)
return -EINVAL;
dev = (struct legacy_ap_device *)calloc(1, sizeof(*dev));
if (!dev)
return -ENOMEM;
dev->device.common.tag = HARDWARE_DEVICE_TAG;
dev->device.common.version = 0;
dev->device.common.module = const_cast<hw_module_t*>(module);
dev->device.common.close = legacy_ap_dev_close;
dev->device.create_audio_policy = create_legacy_ap;
dev->device.destroy_audio_policy = destroy_legacy_ap;
*device = &dev->device.common;
return 0;
}
|
create_audio_policy()中的aps_ops参数指针代表的是,它是AudioPolicyService与外界交互的接口:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
|
struct audio_policy_service_ops aps_ops = {
open_output : aps_open_output,
open_duplicate_output : aps_open_dup_output,
close_output : aps_close_output,
suspend_output : aps_suspend_output,
restore_output : aps_restore_output,
open_input : aps_open_input,
close_input : aps_close_input,
set_stream_volume : aps_set_stream_volume,
set_stream_output : aps_set_stream_output,
set_parameters : aps_set_parameters,
get_parameters : aps_get_parameters,
start_tone : aps_start_tone,
stop_tone : aps_stop_tone,
set_voice_volume : aps_set_voice_volume,
move_effects : aps_move_effects,
load_hw_module : aps_load_hw_module,
open_output_on_module : aps_open_output_on_module,
open_input_on_module : aps_open_input_on_module,
};
|
知道了这些,接下来看create_audio_policy():
create_audio_policy()这个函数作用是创建一个用户自定义的policy_hal模块的接口,因为我们使用的是qcom的芯片,qcom有自己的一套,
android原生有自己的一套,就依照原生的来看吧;其实都是差不多的;
刚刚上面分析的legacy_ap_dev_open()函数有这样一句:
|
......
dev->device.create_audio_policy = create_legacy_ap;
......
|
那这样我们就来看看其create_legacy_ap()函数吧;我们只需要关注的是其中的那么几小段:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
|
static int create_legacy_ap(const struct audio_policy_device *device,
struct audio_policy_service_ops *aps_ops,
void *service,
struct audio_policy **ap)
{
struct legacy_audio_policy *lap;
......
lap = (struct legacy_audio_policy *)calloc(1, sizeof(*lap));
......
lap->policy.set_force_use = ap_set_force_use;
......
lap->service = service;
lap->aps_ops = aps_ops;
lap->service_client =
new AudioPolicyCompatClient(aps_ops, service);
......
lap->apm = createAudioPolicyManager(lap->service_client);
......
*ap = &lap->policy;
......
}
|
就这样,AudioPolicyService.cpp的set_force_use()函数就调用到了这里:
|
audio_policy_hal.cpp
/* force using a specific device category for the specified usage */
static void ap_set_force_use(struct audio_policy *pol,
audio_policy_force_use_t usage,
audio_policy_forced_cfg_t config)
{
struct legacy_audio_policy *lap = to_lap(pol);
lap->apm->setForceUse((AudioSystem::force_use)usage,
(AudioSystem::forced_config)config);
}
|
从之前的create_legacy_ap()函数我们知道apm的由来,
|
lap->apm = createAudioPolicyManager(lap->service_client);
|
createAudioPolicyManager()函数定义在AudioPolicyInterface.h接口中;
|
extern "C" AudioPolicyInterface* createAudioPolicyManager(AudioPolicyClientInterface *clientInterface);
|
而这个createAudioPolicyManager()由硬件厂商实现,返回其AudioPolicyManager;qcom的实现是在AudioPolicyManagerALSA.cpp中;
再往下不做具体分析了,主要是根据不同的策略来切换不同的Output和input设备以及其他一些操作;如果想进一步分析的话,还需要关注
AudioPolicyManagerBase.cpp;
其实准确的总结起来是AudioPolicyService是一个壳子,这个壳子的重要关键就是audio_policy,真正的实现可以由厂商来自己实现,当然
android也有,就是AudioPolicyManagerDefault;
现在回到我的bug中来,bug是这样的,连接蓝牙耳机,在通话过程中,选择切换到Speaker时,output切换到了Earpiece,再进行一次切换,
才会成功切换到Speaker,从下面的log可以看出来,audio path最后是切换到了0x1,而且中间也切到了0x2,又被谁给切到了0x1,
这里是关键;(0x1代表Earpiece,0x2则代表Speaker,0x20代表的是蓝牙耳机);
1
2
3
4
5
6
7
8
9
10
11
12
|
01-02 10:50:26.355 D/ALSADevice( 307): route: devices 0x1 in mode 2
01-02 10:50:26.605 D/ALSADevice( 307): route: devices 0x1 in mode 2
01-02 10:50:27.306 D/ALSADevice( 307): route: devices 0x20 in mode 2
01-02 10:50:27.416 D/ALSADevice( 307): route: devices 0x20 in mode 2
01-02 10:50:27.436 D/ALSADevice( 307): route: devices 0x20 in mode 2
01-02 10:50:27.446 D/ALSADevice( 307): route: devices 0x20 in mode 2
01-02 10:50:33.262 D/ALSADevice( 307): route: devices 0x2 in mode 2
01-02 10:50:33.442 D/ALSADevice( 307): route: devices 0x2 in mode 2
01-02 10:50:33.582 D/ALSADevice( 307): route: devices 0x1 in mode 2
01-02 10:50:33.883 D/ALSADevice( 307): route: devices 0x1 in mode 2
01-02 10:50:33.913 D/ALSADevice( 307): route: devices 0x1 in mode 2
01-02 10:50:33.923 D/ALSADevice( 307): route: devices 0x1 in mode 2
|
中间的分析过程就不说了,直接说结果吧;首先,在Bluetooth的app中有一个HeadsetStateMachine.java,它的作用类似于BT,
WIFI的状态机,在我从蓝牙耳机切换到Speaker的时候,它受到了EVENT_TYPE_AUDIO_STATE_CHANGED的状态消息的回馈,
消息接收到以后,其case走到HeadsetHalConstants.AUDIO_STATE_DISCONNECTED:,因为这时候蓝牙耳机断开;
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
|
case HeadsetHalConstants.AUDIO_STATE_DISCONNECTED:
if (mAudioState != BluetoothHeadset.STATE_AUDIO_DISCONNECTED) {
mAudioState = BluetoothHeadset.STATE_AUDIO_DISCONNECTED;
mAudioManager.setBluetoothScoOn(false);
if (mA2dpSuspend) {
if ((!isInCall()) && (mPhoneState.getNumber().isEmpty())) {
log("Audio is closed,Set A2dpSuspended=false");
mAudioManager.setParameters("A2dpSuspended=false");
mA2dpSuspend = false;
}
}
broadcastAudioState(device, BluetoothHeadset.STATE_AUDIO_DISCONNECTED,
BluetoothHeadset.STATE_AUDIO_CONNECTED);
}
transitionTo(mConnected);
|
最悲惨的在于mAudioManager.setBluetoothScoOn(false);和mAudioManager.setSpeakerphoneOn(true);一样也是强制切换audio Path,
所以这里的setBluetoothScoOn(false)会把audio path的输出设备切换到NONE的状态,底层的处理是在FOR_COMMUNICATION情况下,
NONE会切到Earpiece;所以才会导致第一次切换会到Speaker上的情况,而第二次切换并不会收到
EVENT_TYPE_AUDIO_STATE_CHANGED的状态消息,也就会正常切换到Speaker,所以这里需要修改成:
|
if(mAudioManager.isSpeakerphoneOn()){
mAudioManager.setBluetoothScoOn(false);
mAudioManager.setSpeakerphoneOn(true);
} else {
mAudioManager.setBluetoothScoOn(false);
}
|
判断SpeakerOn的状态,如果为true,则在关闭Bluetooth SCO通路后,再打开Speaker,这样就会重新又会切换到Speaker,测试成功;
最后再说一下,在分析这个bug的时候在stackoverflow上搜索到的一个问题解决办法把我拉进了深巷,当然,其不是这个bug的解决办法,
但是对于有些机器setSpeakerphoneOn调用没有作用的问题可能是一个解答;
http://stackoverflow.com/questions/12036221/how-to-turn-speaker-on-off-programatically-in-android-4-0
1
2
3
4
5
6
7
8
9
10
11
12
|
if setSpeakerOn(true) is not useful;
Class audioSystemClass = Class.forName("android.media.AudioSystem");
Method setForceUse = audioSystemClass.getMethod("setForceUse", int.class, int.class);
// First 1 == FOR_MEDIA, second 1 == FORCE_SPEAKER. To go back to the default
// behavior, use FORCE_NONE (0).
setForceUse.invoke(null, 1, 1);
The combination FOR_MEDIA, FORCE_SPEAKER is typically only
used internally to route the FM-radio audio to the loudspeaker (since
the FM-radio requires you to have a wired headset / headphone plugged in
to act as an antenna). Devices that don't have FM-radio functionality
(or uses an alternative implementation) might ignore this combination of
parameters, so this method would not work on such a device.
|
所有评论(0)